SAGE (Literature Search) - Short-Form (Computer Science): leaderboard
Metric: Exact match (%, 0-100): the target paper appears in the answer text or its citations; the 150 computer science short-form questions of SAGE, each built from a seed paper's metadata, figure and table details and reference overlap with related papers, with one ground-truth paper; deep research agents with their native web search: GPT-5 series through the API at medium reasoning effort with web search enabled, Gemini 2.5 with dynamic thinking and web search, the open DR Tulu agent served with vLLM; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 57.3 | #91 |
| 2 | GPT-5 Mini | 40 | #176 |
| 3 | dr-tulu | 36 | #188 |
| 4 | GPT-5 Nano | 30.7 | #415 |
| 5 | Gemini 2.5 Flash | 30.3 | #237 |
| 6 | Gemini 2.5 Pro | 27.7 | #145 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=sage-literature-search-short-form-computer-science · How It Works · Data refreshed daily, snapshot 2026-10-11.